Laboratory method for predicting calcium sulfate scaling thresholds

Technol. , 1967, 1 (1), pp 41–45. DOI: 10.1021/es60001a005. Publication Date: January 1967 ... Environmental Science & Technology 1967 1 (10), 815-8...
0 downloads 0 Views 5MB Size
CURRENT RESEARCH Laboratory Method for Predicting Calcium Sulfate Scaling Thresholds Julius Glater, Louis Ssutu, and Joseph W. McCutchan Department of Engineering, University of California, Los Angeles, Calif.

An experimental method for measuring calcium sulfate scaling thresholds of natural saline water samples at 100" C. is based on visual observation of freshly formed hemihydrate needles from evaporating brine solution. A novel all-glass laboratory evaporator has been developed. The crystal phase was identified as pure calcium sulfate hemihydrate by x-ray diffraction. Critical concentration factors measured for sea water. Roswell brackish water, and Salton Sea water were 3.2, 2.6: and 1.1, respectively. The influence of ionic strength on calcium sulfate solubility is clearly indicated. A graphical method relates scaling threshold to calcium and sulfate ion concentration for a given saline water. This technique should be useful in estimating the calcium sulfate scaling potential of any natural water in a distillation process.

he deposition of mineral scale from evaporating brine solutions poses a serious problem in the development of reliable and economical equipment for saline water distillation. Technology of alkaline scale control has been substantially advanced during the past few years, but calcium sulfate deposition remains as a potential problem. Alkaline scale, consisting of calcium carbonate and magnesium hydroxide, is formed by reactions resulting from the thermal breakdown of bicarbonate ion which occurs in all natural saline waters. Bicarbonate ion can be effectively decomposed by injecting a low concentration of strong acid into evaporator feed water. Acid injection followed by deaeration has been used successfully for control of alkaline scale in many large distilling plants (Checkovich, 1964; Mulford er al., 1965). Calcium sulfate scale is precipitated when the equilibrium solubility of a given crystalline form of this compound is exceeded. This occurs when certain saline waters are concentrated by distillation. Although anhydrite (CaS04) is the most stable crystalline modification at usual evaporator temperatures, the somewhat more soluble hemihydrate ( C a S 0 4 . */,H,O) precipitates first and is found most frequently in actual scale samples, because of the long time required (approximately 17 hours at 100" C.) before any hemihydrate converts to the more stable anhydrite form. That the solubility of calcium sulfate is also strongly influenced by temperature can be seen from the steeply inverted solubility-temperature relationship of these two scale modifications. I n distillation processes we are therefore concerned with critical conditions tem -

T

perature and concentration factor) for hemihydrate precipitation from a given saline water. Purpose of Work

We have devised a rapid experimental method for determining concentration thresholds at which calcium sulfate scale begins to deposit from a saline water sample at its atmospheric boiling point. The met!iod is rapid and easily reproducible by inexperienced personnel. I t involves simple laboratory apparatus and small samples of water, and furthermore avoids the necessity of long runs in commercial distilling equipment. Many recent studies o n equilibrium solubility of calcium sulfate appear in the literature (Dickson er al., 1963; Marshall et al., 1964; Power et al., 1964). These reports are based largely on solubility in pure water or solutions of sodium chloride. Solubility data obtained from distilled water or dilute salt solutions cannot be applied directly t o sea water, a complex system of mixed electrolytes. For example, the solubility of certain sparingly soluble salts may be increased as much as three times in 0.1M salt solutions of monovalent type and ten times in 0.1M solutions of divalent type. I t is therefore difficult to predict equilibrium solubilities in sea water (about 0.5M in NaC1) from data obtained from simple solutions. Fundamental work o n hemihydrate solubility in natural sea water has been reported by Langelier et al. (1950) and Toriumi et 01. (1 933). The Langelier study gives a threshold concentration factor of 3.1 for deposition of calcium sulfate scale at 100" C. Calculations by York and Schorle (1966) based o n Toriumi's data agree closely with the Langelier value. Concentration factor (CF) is a frequently misunderstood term in engineering literature. It may be defined as the ratio of salinities of blowdown to feed water for a n evaporator. I n a batch distillation process, concentration factor is defined as follows: CF

=

initial volume of saline water volume of residual brine

Table I shows the relationship between concentration factor and other common units for expressing sea water concentration. Langelier's value of 3.1 has been used widely for calculation of safe blowdown ratios in various types of distilling equipment. Unfortunately, the Langelier work was done in sea water at 60" and 100" C. only. Several important questions o n calcium sulfate precipitation threshold still remain to be anshered. Volume 1, Number 1, January 1967

41

I

Table I. Sea Water Concentration Units Concn. Factor 1 1 1 1 2 2 3 3 4 4 5

0 1 2 5 0 5 0 5 0 5 0

z Distilled Water

0 0 9 0 170 330 500 600 670 715 750 778 800

z

% Brine 0.0 91.0 83.0 67.0 50.0 40.0 33.0 28.5 25.0 22.2 20.0

% Increase Total in Dissolved Concn. Salts 0 10 20 50 100 150 200 250 300 350 400

3 3 4 5 7 8 10 12 14 15 17

50 85 20 25 00 75 50 00 00 75 50

-CHIPS

0 5 0 5 0 5 0 0 0 5 0

Figure I .

1. What is the scaling threshold of sea water samples at different dilutions? For example, how can one experimentally evaluate the scaling potential of waters from bays or coastal regions near rivers ? 2. What is the scaling threshold of natural saline waters with different ionic distributions than sea water? 3. How effective are various water pretreatment procedures for control of calcium sulfate scale? For example, how can one experimentally evaluate the scaling potential of waters treated with precipitating agents, complexing agents, o r ion exchange resins ? 4. What is the scaling threshold of natural waters at distillation temperatures above 100" C.? This work describes an accelerated experimental method for providing answers to questions 1 through 3. By the technique proposed here, it is possible to set design limitations for preventing calcium sulfate scale in an idealized distillation process and to gather data for any saline water sample distilled at 100" C. The method is based on conditions approaching equilibrium between solid hemihydrate and its dissolved ions. I n actual distillation practice, attainment of equilibrium is not always possible and some variation between different types of distilling equipment may be noted. Sea water data collected thus far are in good agreement with Langelier. This series of experiments was carried out at the atmospheric boiling point only. The effect of temperatures above 100" C. o n calcium sulfate solubility in sea water has not been rigorously studied. Hemihydrate curves relating sea water concentration factor with temperatures up to 300" F. have been drawn from limited experimental and operational data (Skerritt and Howe, 1966; Standiford and Sinek, 1961). The exact position of these curves is controversial and provides no clear limit for prediction of scale-free operation in distilling plants. We are currently applying our technique at higher temperature levels in an effort to provide additional data on the solubility of hemihydrate in sea water. Apparatus and Experimental Procedure Boiling water containing dissolved calcium and sulfate ions will immediately deposit hemihydrate crystals upon reaching

42 Environmental Science and Technology

LIQLID LEVEL

CIRBORUNDUM

Salinity 35 38 42 52 70 87 105 120 140 157 175

I

Modified Thiele tube

the saturation point. Observance of this phenomenon suggested the possibility of devising a visual method for detecting the first appearance of hemihydrate crystals in evaporating brine solutions. Natural sea water obtained from Marineland in Los Angeles was first acidified with hydrochloric acid to a methyl orange end point (pH 4.0) in order to prevent the formation of alkaline scale, The acidified water was then evaporated slowly to CF g 2.0. A portion of this preconcentrated water was placed in a small beaker containing a glass immersion heater and evaporation was continued. It was anticipated that crystals would appear on the immersion heater at the point of saturation. As the brine became more concentrated, however, violent bumping was observed. Bumping was not improved with boiling chips but could be avoided by stirring rapidly with a magnetic stirrer. Turbulence in the solution caused crystals to flake off the heat transfer surface as soon as they formed. The resulting slurry of fine needles developed slowly after saturation was achieved. No definite saturation end point could be observed in the rapidly moving solution. Attempts to improve this system were ultimately abandoned in favor of a more efficient glass evaporator. A device was finally developed based on the Thiele-Dennis tube (Dennis, 1920) commonly used for melting point baths in organic chemistry laboratory, modified as shown in Figure 1. The wide neck was added to enhance evaporation from the liquid surface and to provide a surge volume in case of bumping. Carborundum chips sealed into the narrow arm facilitate more even boiling and promote nucleation of calcium sulfate crystals. The Thiele tube configuration enhances natural brine circulation at a fairly slow rate. This geometry together with Carborundum chips at the point of heat input ensures even boiling with no bumping observed. To determine the critical concentration factor for hemihydrate deposition in sea water, the following procedure is followed: About 3 liters of acidified sea water (pH 4.0) is evaporated slowly on a hot plate with stirring to a concentration factor about 0.5 C F unit lower than the anticipated critical concentration factor. The preconcentrated solution is now filtered by first stirring in 50 grams of Celite and transferring the milky suspension to a fine sintered glass vacuum filter.

It is very important to remove small suspended particles from the preconcentrated brine, since this method is based on a visual end point observed under magnification. Even small traces of solid impurity may interfere. The Thiele tube is filled to the indicated mark with a measured volume of brine filtrate. Another portion of filtrate is used to fill a 50-ml. buret. The apparatus is now assembled as shown in Figure 2, with heat applied to the narrow arm through heating tape connected to a Variac. The power input is adjusted to produce gentle boiling and brine circulation. Water vapor is discharged directly into the air as evaporation proceeds, and a constant ..... .~ liquid level In the Thiele tube is maintained by slow addition of fresh filtrate from the buret. When boiling brine becomes saturated with calcium sulfate, a slurry of fine crystal needles suddenly appears in the body of the circulating fluid. These probably form on the Carborundum chips first and are then dislodged by the fast-flowing brine stream in the narrow arm. The end point at which crystals first appear can be observed in a darkened room by using a 1OX hand lens or small telescope. The visual field is illuminated by a high intensity light source placed at about 90" to the line of sight. Observations are made in the wide arm of the tube, where liquid circulation rate is slow. End point detection may he substantially improved by using plane-polarized light in place of ordinary light. Polarized light has been used by Gross (1954) for detection of sodium chloride crystals from biological fluids. Apparatus shown in Figure 3 must be arranged so that telescope, large arm of Thiele tube, and light source are oriented in a straight line. A piece of Polaroid glass is placed between the tube and light source and another piece between the tube and telescope. The two pnlarizers are oriented at 90" to one another, so that a dark field is observed. Polarized light passing through a water sample will he stopped by the polarizer placed at right angles to the plane of polarization. When small crystals appear in the water, light will be refracted from these crystals in such a manner as to alter the plane of polarization. Some of the refracted rays will now pass through the second polarizer. Small crystals appear as bright points of light on a dark field. This modification gives sharper end points and has improved the precision of these measurements. Vapor bubbles do not refract polarized light and appear transparent in this system. At the completion of each run, the Thiele tube is filled with distilled water, heated to the boiling point for a few minutes, then rinsed with fresh distilled water and placed in a glass drying oven overnight. If tubes are not cleaned in this manner, the Carborundum chips become fouled with brine or salt crystals and violent humping will result in subsequent runs.

il

POWER

POWER

Figure 2. Experimental apparatus using ordinary light

Calculations When the end point is reached, power is turned off and the Thiele tube is cooled to room temperature. A buret reading is taken and the volume of added filtrate recorded. The volume of concentrated brine in the Thiele tube is measured by transferring the cool contents to a graduated cylinder. Although the evaporation is carried out at constant liquid volume, there is usually a small discrepancy between the initial and final Thiele tube volumes. Concentration factor may he calculated from the following relationship: VXFp CF = (Vi ~Vf

Figure 3. Experimental apparatus using polarized light

+

Volume 1, Number 1, January 1967 43

Table 11. Analytical Data on Three Saline Water Samples Source of Saline Water

Cat*

SOPS

Concn

Concn.

Sea water, Marineland, Calif.

400 p.p.m. 1.00 X 10-*M/I.

2600 p.p.m. 2.71 X 10-2M/I.

Brackish water, Roswell, N. M.

488 p.p.m. 1.22 X 10-2M/I.

1510 p.p.m. 1.58 X 10-*M/I.

Saline water, Salton Sea, Calif.

616 p . p m 1.54 X 10-2M/1.

7260 p.p.m. 7.56 X 10-2M/I.

t

CaSO,

Ion Product

Critical CF

x

3.18 3.16 3.19

2.71

10-4

1 . 9 3 x lo-'

2.61 2.58 2.60

x

1.10 1.10 1.13

1.17

10-3

I

I

Figure 4. Hemihydrate crystals from a Thiele tube run (8OX magnification)

Vi

= initial volume in Thiele tube V, = volume added from buret CF, = concentration factor of preconcentrated water V I = final volume in Thiele tube

Results and Di.scussion Figure 4 is a photomicrograph of crystals obtained from a typical run on sea water. The needles are hexagonal prisms characteristic of hemihydrate. A sample of this material was purified, dried, and analyzed by x-ray diffraction at the UCLA Department of Geophysics. The crystals were positively identified as pure calcium sulfate hemihydrate. Data collected thus far on three natural saline waters are presented with analytical data in Table 11. The average value obtained for Marineland sea water is approximately 0.1 C F unit higher than the 3.1 figure obtained by Langelier. There 44

Environmental Science and Technolog)

are two possible explanations for this discrepancy. Composition of the two water samples may vary slightly or there may be a small time lag between saturation point (point at which crystals deposit on Carbonmdum chips) and the point when crystals are first observed suspended in the brine. This lag would cause more water to evaporate, resulting in a slightly larger calculated concentration factor. It would be desirable to determine the true equilibrium solubility of hemihydrate in each water sample and compare this figure with Thiele tube values. An interesting conclusion may be derived from data presented in Table 11. The critical concentration factor for Roswell water is 2.6 compared with 3.2 for Marineland sea water. These data would predict a potentially more serious calcium sulfate scaling problem at the Roswell, N. M., vapor compression distilling plant than a similar plant operating with sea

water. Applying the solubility product principle without regard to salt effect o r activity coefficients, we can calculate an ion product (IP) for calcium sulfate in each of the above-mentioned waters. Marineland sea water I P

= =

Roswell water IP

= =

[Ca”] [S04-2] (0.0100) (0.0271) [Ca“’] [SO4-O-] (0.0122) (0.0158)

=

2.71 X lop4

=

1.93 X IOp4

These values are of the same order of magnitude and show that both waters would be supersaturated with respect to calcium sulfate (K,,, = 2.4 X 10-j) if other dissolved ions were not present. From these calculations one would also expect approximately the same critical CF value for both waters. In fact, on the basis of these calculations, sea water should deposit calcium sulfate at a lower concentration factor than Roswell water. Since neither of these conditions is met, it is evident that calcium sulfate solubility is strongly influenced by ionic strength of the solution in which it is dissolved. Sea water with a salinity of 35 presents a less serious scaling problem than Roswell water with a salinity of only 16, even though the calcium sulfate ion product is larger for sea water. The profound effect of background ions on calcium sulfate solubility is clearly demonstrated by this situation. We are concerned not only with total salinity but also with relative distribution and valence type of individual ions. Sea water, for example, has a magnesium ion concentration about four times that of Roswell water. Magnesium and certain other divalent cations have a large positive effect on solubility of sparingly soluble salts. Estimation of solubility from analytical data alone would appear to be a monumental task. The authors believe that scaling potential of any water sample can be most readily determined from good equilibrium solubility measurements or by the method presented in this paper. One further application of tlie Thiele tube technique might be considered now. Assume that a certain type of water is treated by ion exchange (McIlhenny, 1965) t o remove either C a f 2 o r SodF2 but not appreciably alter the distribution of background ions, Under these circumstances, one would anticipate a higher critical concentration factor. Using Marineland sea water, for example, we know that saturation occurs at 100” C. when [Ca+’] and [S04-2] are about 3.2 times their natural concentration. The following equation may therefore be written for a corrected solubility product constant K*,, of hemihydrate in sea water:

3.2[Caf21 X 3.2

=

K*,,,

From Marineland water analysis, converting parts per million to molarity, we may write K*,,

=

(3.2)2(0.010) (0.027)

=

2.75

>(

We can now set up a general equation for the solubility product of calcium sulfate in terms of the critical concentration factor (CF)2 [CaA2]

=

K*$,, = 2.75 X 10-3

This equation describes the boundary condition for hemihydrate precipitation from sea water at any calcium and sul-

fate concentration. By substituting arbitrary values for [Caf21 [S04-2], we can solve for C F , using the following relationship:

The family of curves shown in Figure 5 was derived by substituting values for calcium and sulfate above and below normal sea water levels. Some experimental verification for the calculated curves has been obtained and data points are shown as circles o n the diagram. These data were collected by first stripping Marineland sea water of all calcium and sulfate. Calcium ion was removed by adding a stoichiometric quantity of sodium oxalate, and sulfate ion by adding a stoichiometric quantity of barium chloride. Measured amounts of calcium chloride and sodium sulfate were then added to this depleted water and samples were run by the Thiele tube technique. Figure 5 could be developed for any natural water and would be useful in predicting scaling thresholds if the critical C F value and analytical data on calcium and sulfate content were available and especially would be useful in evaluating the effectiveness of ion exchange softening for prevention of calcium sulfate scale. Acknowledyinent

The authors gratefully acknowledge the conscientious efforts of K. Tasugi, K. Fung, and C. Keehn throughout this study. They are also indebted to D. J. Albright for his able glassblowing services and to E. S. C. Bowler for photomicrography. Literature Cited

Checkovich, A., U. S. Patent 3,119,752 (Jan. 28, 1964). Dennis, L. M., J . Ind. Eng. Chem. 12, 366 (1920). Dickson, F. W., Blount, C. W., Tunell, G., Am. J . Sci. 261, 61 (1963). Gross, W. J., J . Exprl. B i d . 31, 402 (1954). Langelier, W. F., Caldwell, D. H., Lawrence, W. B., Znd. Eng. Cliem. 42, 126 (1950). Marshall, W. L., Slusher, R., Jones, E. V., J . Ckem. Eng. Data 9, 187 (1964). McIlhenny, W. F., Proceedings of First International Symposium on Water Desalination, Washington, D. C., October 1965. Mulford, S. F., Glater, J., McCutchan, J. W., “Proceedings of First International Symposium on Water Desalination,’’ Washington, D. C., October 1965. Power, W. H., Fabuss, B. M., Satterfield, C. N., J , Chem. Eng. Datu 9, 437 (1964). Skerritt, D. E., Howe, E. D., University of California, Berkeley, Rept. 66-1, 7-9 (January 1966). Standiford, F. C., Sinek, J. R., Chem. Eng. Progr. 57, 59 (1961). Toriumi, T., Kuwahara, T., Hara, R., Kogjo Kuguku Zrtsslii 36, 1651 (1933). York, L. J., Schorle, B. J., “Principles of Desalination,” K. S. Spiegler, ed., Chap. 10, Academic Press, New York,1966. Receiced for reciew October 6 , 1966. Accepted Deceniber I , 1966. Work supported by saline water research ,funds procided b y the California State Legislature as u speciul item in the Unicersity of California budgef. Volume 1, Number 1, January 1967 45